Harvard Law Review Harvard Law Review Harvard Law Review

Article II

“As the Legislature Has Prescribed”: Removing Presidential Elections from the Anderson-Burdick Framework

1×1. Bush v. Gore, 531 U.S. 98, 104 (2000) (percuriam) (“When the state legislature vests the right to vote for President in its people, the right to vote as the legislature has prescribed is fundamental….”)
In the summer of 2019, California enacted Senate Bill 27 (S.B. 27), or the Presidential Tax Transparency and Accountability Act.2×2. Cal. Elec. Code §§ 6880–6884 (West 2019), invalidated in part by Patterson v. Padilla, 451 P.3d 1171 (Cal. 2019); id. §§ 8900–8903 (amended 2021); Nick Cahill, Trump Tax Returns Required by New California Law, Courthouse News Serv. (July 30, 2019), https://www.courthousenews.com/trump-tax-returns-required-by-new-california-law [https://perma.cc/EVJ7-SMTM]. The law required candidates for President and Governor wishing to appear on the state’s primary ballot to file the candidate’s five most recent income tax returns with the California Secretary of State, who would then publish redacted versions.3×3. Cal. Elec. Code §§ 6883–6884 (presidential candidates), invalidated by Patterson, 451 P.3d 1171; id. §§ 8902–8903 (gubernatorial candidates) (amended 2021). The fact that the law applied only to primary ballots went curiously unexplored in the ensuing media discussion and litigation. While the California Constitution requires candidates for state and congressional office to have appeared on the primary ballot in order to appear on the general ballot, see Cal. Const. art. II, § 5(a)–(b), it does not require the same of presidential candidates, see id. § 5(c)–(d). Though neutral in its language, the law was widely perceived as inspired by President Trump, who had declined to release his tax returns prior to both the 2016 and 2020 elections, even though presidential candidates had consistently done so in the preceding decades.4×4. Joseph J. Thorndike, Tax History: From Nixon to Trump: A Short History of Voluntary Tax Disclosure, 102 Tax Notes 612, 614 (2019). Indeed, while California was the only state to pass such a law, the legislatures of other blue states had considered similar measures.5×5. Donna Borak, State Lawmakers Move to Require Tax Returns from Presidential Candidates — Including Trump, CNN (Apr. 24, 2019, 7:01 AM), https://www.cnn.com/2019/04/24/politics/presidential-tax-returns-states-2020-trump/index.html [https://perma.cc/YT77-YU53]. Democrats in New York had even titled their proposal the Tax Reports Uniformly Made Public (TRUMP) Act. Chris Isidore, TRUMP Act Would Put Trump’s New York Tax Returns Online, CNN (May 5, 2017, 2:41 PM), https://money.cnn.com/2017/05/05/news/economy/new-york-trump-tax-returns/index.html [https://perma.cc/PC5X-RWK7].

The constitutionality of California’s law was immediately subject to debate. Pro-Trump Republicans, predictably, insisted that the law was unconstitutional and brought suit in both state and federal court to have it struck down.6×6. California Republican Party Files Two Simultaneous Lawsuits over SB 27: Tax Transparency Bill — The New Law Requires Candidates for U.S. President and California Governor to Disclose Their Income Tax Returns, Sierra Sun Times (Aug. 6, 2019, 12:38 PM), https://goldrushcam.com/sierrasuntimes/index.php/news/local-news/19661-california-republican-party-files-two-simultaneous-lawsuits-over-sb-27-tax-transparency-bill-the-new-law-requires-candidates-for-u-s-president-and-california-governor-to-disclose-their-income-tax-returns [https://perma.cc/JX3E-YFEX]. However, several legal luminaries came to the law’s defense. Governor Gavin Newsom’s signing statement was accompanied by brief arguments in favor of the law’s constitutionality from Professor Erwin Chemerinsky and prominent attorneys Theodore Boutrous, Jr., and David Boies.7×7. Press Release, Office of Governor Gavin Newsom, Governor Gavin Newsom Signs SB 27: Tax Transparency Bill (July 30, 2019), https://www.gov.ca.gov/2019/07/30/governor-gavin-newsom-signs-sb-27-tax-transparency-bill [https://perma.cc/7EU9-4XDM]. Chemerinsky later expanded his position into an op-ed. See Erwin Chemerinsky, Opinion, California’s New Law Requiring Presidential Candidates to Disclose Tax Returns Is Constitutional, L.A. Times (July 31, 2019, 10:16 AM) https://www.latimes.com/opinion/story/2019-07-31/california-law-candidates-tax-returns-constitutional [https://perma.cc/SJ2W-VJ79]. Other prominent law professors weighed in on both sides of the debate.8×8. See, e.g., Jerry Lambe, Harvard Law Profs Clash over California Law Aimed at Trump’s Tax Returns, Law & Crime (July 31, 2019, 11:49 AM), https://lawandcrime.com/high-profile/harvard-law-profs-clash-over-california-law-aimed-at-trumps-tax-returns [https://perma.cc/JW23-NDAW].

Courts proved to be more in agreement with the law’s detractors. Roughly two months after the law’s enactment, Judge England of the Eastern District of California granted a preliminary injunction barring the enforcement of the law against presidential candidates.9×9. Griffin v. Padilla, 417 F. Supp. 3d 1291, 1308 (E.D. Cal. 2019). In determining whether the plaintiffs were likely to succeed on the merits, Judge England found the law wanting. Most notably for present purposes,10×10. In addition, Judge England concluded that the Ethics in Government Act of 1978 (EIGA), Pub. L. No. 95-521, 92 Stat. 1824 (codified as amended in scattered sections of 2, 5, and 28 U.S.C.), which also requires presidential candidates to make financial disclosures (but not to release tax returns), preempted California’s law. See Griffin, 417 F. Supp. 3d at 1306–07. He also cited U.S. Term Limits, Inc. v. Thornton, 514 U.S. 779 (1995), to conclude that the law violated the Presidential Qualifications Clause by creating new qualifications for the presidency. See Griffin, 417 F. Supp. 3d at 1298–302. U.S. Term Limits itself actually considered only the Legislative Qualifications Clauses. See U.S. Term Limits, 514 U.S. at 782–83. However, the case’s logic is largely applicable to presidential elections. See Derek T. Muller, Weaponizing the Ballot, 48 Fla. St. U. L. Rev. 61, 87 (2020). Judge England applied the AndersonBurdick11×11. Anderson v. Celebrezze, 460 U.S. 780 (1983); Burdick v. Takushi, 504 U.S. 428 (1992). balancing test, concluding that the state’s asserted interests could not justify such an egregious infringement on voters’ First Amendment rights as the framework construed them.12×12. See Griffin, 417 F. Supp. 3d at 1303–05. Applying the same framework, he also determined that the law violated the Equal Protection Clause because it did not apply to independent candidates, who do not participate in presidential primaries.13×13. See id. at 1305–06. California appealed to the Ninth Circuit, but before that court could rule, it dismissed the case as mooted by the decision of the California Supreme Court that the law’s application to presidential elections violated the California Constitution.14×14. Griffin v. Padilla, Nos. 19-17000 et al., 2019 WL 7557783, at *1 (9th Cir. Dec. 16, 2019) (citing Patterson v. Padilla, 451 P.3d 1171, 1191 (Cal. 2019)). The state constitution’s mandate that the Secretary of State list those “recognized . . . throughout the nation or throughout California” as presidential candidates on the primary ballot was not compatible with the disclosure requirement. Patterson, 451 P.3d at 1173 (quoting Cal. Const. art. 2, § 5(c)).

Thus, the issue was resolved without any definitive decision as to whether California’s law violated the Federal Constitution. But it is only a matter of time before the question comes up again. As noted above, laws similar to California’s had been proposed in other states, and there is no reason that laws of this type may be used only to seek tax returns. In 2011, Arizona Governor Jan Brewer vetoed a bill that would have required presidential candidates to present proof of citizenship to appear on that state’s ballot — the state legislature had passed the bill after the rise of the “birther” conspiracy theory that President Obama is not a natural-born citizen.15×15. Arizona Governor Jan Brewer Vetoes “Birther” Bill, BBC (Apr. 19, 2011), https://www.bbc.com/news/world-us-canada-13125337 [https://perma.cc/57SR-E4ZD].

As the law stands today, state balloting requirements make it difficult for third-party candidates to gain ballot access.16×16. See Quinn Scanlan, Coronavirus Adds to Ballot Access Hurdles for Third Party Candidates, ABC News (May 23, 2020, 9:39 AM), https://abcnews.go.com/Politics/coronavirus-adds-ballot-access-hurdles-3rd-party-candidates/story?id=70816388 [https://perma.cc/FDH3-C2NU]. And, as President Trump was quick to point out,17×17. Full Transcript: First 2016 Presidential Debate, Politico (Sept. 27, 2016, 1:55 AM), https://www.politico.com/story/2016/09/full-transcript-first-2016-presidential-debate-228761 [https://perma.cc/85ZD-ETVU]. presidential candidates are required to make financial and other disclosures under the Ethics in Government Act of 197818×18. Pub. L. No. 95-521, 92 Stat. 1824 (codified as amended in scattered sections of 2, 5, and 28 U.S.C.). (EIGA). But at the same time, it would seem undemocratic for a state to pass a law denying presidential ballot access to “any member of the Republican Party” or to “Joseph R. Biden, Jr.,” or, perhaps, to pass a neutrally worded law calculated to accomplish the same thing. And of course, there are countless ways to ensure an electoral outcome without altering ballot access laws, such as by making it easier or harder for certain demographic groups to cast presidential ballots.19×19. Cf. Obama for Am. v. Husted, 697 F.3d 423, 425 (6th Cir. 2012) (invalidating the decision of the Ohio Secretary of State to allow additional early in-person voting for military voters).

The Supreme Court has provided many guideposts to navigate this difficulty, the most important of which is the Anderson-Burdick balancing test. The test takes its name from Anderson v. Celebrezze,20×20. 460 U.S. 780 (1983). a 1983 decision striking down an Ohio law that placed an unjustified burden on independent candidates seeking access to the presidential ballot,21×21. Id. at 806. and Burdick v. Takushi,22×22. 504 U.S. 428 (1992). a 1992 decision upholding Hawaii’s prohibition on write-in voting.23×23. Id. at 441. The two cases are still regularly cited together in decisions passing on electoral restrictions of all kinds.24×24. See, e.g., Republican Nat’l Comm. v. Democratic Nat’l Comm., 140 S. Ct. 1205, 1209 (2020) (Ginsburg, J., dissenting). The balancing test requires courts to weigh the injury to an individual’s First and Fourteenth Amendment rights, left undefined, against the state’s interest in imposing a given election regulation.25×25. See, e.g., Griffin v. Padilla, 417 F. Supp. 3d 1291, 1303 (E.D. Cal. 2019).

Judge England’s conclusion that the California law failed to pass muster under this test is likely correct. The exclusion of major political candidates on grounds of insufficient paperwork is a hallmark of weaker democracies,26×26. See, e.g., Benin Holds Vote with No Opposition Candidates, BBC (Apr. 28, 2019), https://www.bbc.com/news/world-africa-48084124 [https://perma.cc/3WG2-M3FE]; Andrew Roth, Belarus Bans Two Opposition Candidates from Running in Elections, The Guardian (Aug. 2, 2020, 12:55 PM), https://www.theguardian.com/world/2020/jul/14/belarus-bans-two-opposition-candidates-from-running-in-elections [https://perma.cc/VB26-F7D3]. and it is reasonable to conclude that the burden posed by such a restriction on California’s voters could not be justified by somewhat greater financial transparency.27×27. There is reason to doubt that President Trump would have complied with the law’s requirements, given that California was not important to his reelection strategy. See, e.g., Vikram David Amar, Can and Should States Mandate Tax Return Disclosure as a Condition for Presidential Candidates to Appear on the Ballot?, Justia: Verdict (Dec. 30, 2016), https://verdict.justia.com/2016/12/30/can-states-mandate-tax-return-disclosure-condition-presidential-candidates-appear-ballot [https://perma.cc/A5K6-Y68Z]. It presumably mattered even less to his primary strategy. Thus, as long as Anderson-Burdick prevails, states will be limited in their ability to enact election restrictions as long as it can be argued that the value of those restrictions does not outweigh their costs for democracy.

However, the case highlights some of the shortcomings of current law. Practically, the Anderson-Burdick balancing test is frustratingly vague. It does not guide courts in determining what constitutes a constitutional injury or a compelling justification, with the result that, especially in the emotional context of a presidential election, respected legal minds can reach opposite conclusions.28×28. See Lambe, supra note 8. But the more damning flaw in Anderson-Burdick is conceptual: it does not recognize that, as a matter of constitutional theory and actual practice, presidential elections, more than other elections, are left to the discretion of the states.

By returning to the Constitution’s text and revising presidential election caselaw to better accord with it, the Court may be able to solve these problems. This Note argues that a better approach lies in deference to the constitutional authority of states to make their own decisions with regard to presidential elections. Echoing the Supreme Court’s 1892 decision in McPherson v. Blacker29×29. 146 U.S. 1 (1892). and then–Justice Rehnquist’s dissent in Anderson,30×30. Anderson v. Celebrezze, 460 U.S. 780, 806–23 (1983) (Rehnquist, J., dissenting). this deferential approach would strike down state restrictions only when they contravene a separate provision of the Constitution, with those provisions being read in the context of a system where the President is not popularly elected. Thus, while provisions like the First Amendment or the Presidential Qualifications Clause might still limit state authority, those limits would not create a general-purpose balancing test. In addition to ensuring greater textual fidelity, this approach would reduce uncertainty — there would be little doubt that most restrictions implemented by states were constitutional.

Anderson-Burdick applies to all restrictions on all elections, and that one-size-fits-all approach is one of the framework’s key shortcomings.31×31. A generalized critique of the framework can be found in the opinions of Judge Readler on the Sixth Circuit. See, e.g., Daunt v. Benson, 956 F.3d 396, 423–25 (6th Cir. 2020) (Readler, J., concurring in the judgment) (describing Anderson-Burdick as a “dangerous tool” that “affords far too much discretion to judges,” id. at 424). But this Note focuses specifically on presidential elections, to which the framework is particularly ill-suited. If the Supreme Court chooses to strike the test down, the different constitutional provisions governing congressional elections would not require the same level of deference to states. Judicial opinions often assume the existence of a right to vote, but that right need not manifest equally in all elections.32×32. The distinction would not be unusual. For instance, freedom of speech is analyzed differently in the context of commercial speech. See Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n, 447 U.S. 557, 563 (1980). Ballot access restrictions help frame this distinction because they are easily differentiated even in simultaneous elections.33×33. In contrast, it might prove difficult in practice for states to apply different sets of rules in congressional and presidential elections. Cf. Michael T. Morley, Essay, Dismantling the Unitary Electoral System? Uncooperative Federalism in State and Local Elections, 111 Nw. U. L. Rev. Online 103, 113–18 (2017) (exploring a similar issue regarding state and federal elections). But the principles discussed below would be equally applicable to state laws, like voter ID laws or laws regulating polling places, that exclusively governed the casting of presidential ballots. Meanwhile, this Note does not address the question of who within state governments has the authority to control election rules — it does not require acceptance of the “independent state legislature” theory that the state legislature has exclusive authority.34×34. This theory has received both support, see, e.g., Michael T. Morley, The Independent State Legislature Doctrine, Federal Elections, and State Constitutions, 55 Ga. L. Rev. 1, 32–37 (2020), and strong condemnation, see, e.g., Jane Mayer, The Big Money Behind the Big Lie, New Yorker (Aug. 2, 2021), https://www.newyorker.com/magazine/2021/08/09/the-big-money-behind-the-big-lie [https://perma.cc/F9P5-DDEN] (quoting Professor Nathaniel Persily as arguing that the independent legislature doctrine is “giving intellectual respectability to an otherwise insane, anti-democratic argument”). For present purposes, it is sufficient to recognize that the independent state legislature doctrine does not now hold sway — otherwise, the California Supreme Court’s decision striking down S.B. 27 would have been ineffective, as would have been Governor Jerry Brown’s veto of an earlier version of the bill. See Letter from Gov. Edmund G. Brown, Jr., to Members of the California State Senate (Oct. 15, 2017), https://www.ca.gov/archive/gov39/wp-content/uploads/2017/11/SB_149_Veto_Message_2017.pdf [https://perma.cc/DG4P-YCCU].

Part I reviews the constitutional basis for state authority in presidential elections, noting how states have historically exercised this power. Part II reviews Supreme Court decisions that have articulated restraints on the ability of states to manage their presidential elections. Part III argues that many of these decisions are inconsistent with the Constitution, and suggests an alternative approach based on deference to states. Part IV briefly surveys constitutional constraints on state authority that would remain under this approach. This Note concludes that, though a hands-off approach carries risks for democracy, it is nonetheless required as a matter of textual fidelity.

I. Constitutional Underpinnings and Historical Practice

The U.S. President and Vice President are chosen not by popular vote but by the votes of electors designated by individual states. Article II, section 1 of the Constitution provides that “Each State shall appoint, in such Manner as the Legislature thereof may direct, a Number of Electors” equal to the number of the state’s congressional representatives.35×35. U.S. Const. art. II, § 1, cl. 2. This language tracks similar language governing the appointment of delegates under the Articles of Confederation, Articles of Confederation of 1781, art. V, para. 1, under which eleven out of thirteen states had selected delegates without popular input, see Bradley A. Smith & Daniel P. Tokaji, Article I, Section 2, Nat’l Const. Ctr., https://constitutioncenter.org/interactive-constitution/interpretation/article-i/clauses/762 [https://perma.cc/NA2P-BBZD]. The section goes on to give Congress the power to fix the “Time of chusing the Electors” and a single day on which they shall cast their votes.36×36. U.S. Const. art. II, § 1, cl. 4. It also includes the Presidential Qualifications Clause: “No Person except a natural born Citizen” shall be eligible, “neither shall any Person” under the age of thirty-five or having not been a resident of the United States for at least fourteen years.37×37. Id. art. II, § 1, cl. 5. There is an exception to the citizenship requirement for individuals who were U.S. citizens at the time of the Constitution’s enactment. Id. Neither the Twelfth Amendment, which alters the voting procedures of the Electoral College,38×38. Id. amend. XII. nor the Twentieth Amendment, which alters the start date of presidential and vice presidential terms,39×39. Id. amend. XX. affect these provisions — though several constitutional amendments to reduce the discretion of the states were proposed in the nineteenth century and defeated.40×40. See McPherson v. Blacker, 146 U.S. 1, 33–34 (1892). There are, however, constitutional provisions barring states from limiting the right to vote on bases including race,41×41. U.S. Const. amend. XV. sex,42×42. Id. amend. XIX. and age;43×43. Id. amend. XXVI. the Twenty-Fourth Amendment states that poll taxes may not be used to abridge the right to vote “for electors for President or Vice President.”44×44. Id. amend. XXIV.

This broad grant of authority to the states stands in contrast to the constitutional provisions governing the selection of members of the House and Senate.45×45. This contrastive reading largely tracks that of Professor Michael Morley. See Morley, supra note 33, at 108–09. Article I, section 2 provides that members of the House shall be “chosen . . . by the People of the several States” and that voters shall be eligible to participate if they are eligible to vote for “the most numerous Branch of the State Legislature.”46×46. U.S. Const. art. I, § 2, cl. 1. Section 3, on the other hand, specifies that senators are “chosen by the Legislature” of each state,47×47. Id. art. I, § 3, cl. 1. but the Seventeenth Amendment overrides this provision and supplies language mirroring the provision in section 2.48×48. See id. amend. XVII. Section 4 gives state legislatures the power to prescribe the “Times, Places and Manner” of holding congressional elections, but absent the large caveat in the corresponding provision in Article II: “Congress may at any time by Law make or alter” the state regulations on time and manner.49×49. Id. art. I, § 4, cl. 1. Meanwhile, the qualifications clauses for each house parallel that for the President, though with less demanding age and citizenship requirements.50×50. See id. art. I, § 2, cl. 2; id. art. I, § 3, cl. 3.

Especially when read in contrast to the legislative provisions, the phrase “in such Manner as the Legislature thereof may direct” would seem to convey virtually unlimited authority to the states in determining how presidential electors are chosen, regardless of the wishes of the voting public. And indeed, at the time of the Founding, many states did not hold presidential elections at all. In five of the eleven states that cast electoral votes for President Washington in 1789, the state legislatures simply chose the electors themselves.51×51. See James W. Ceaser & Jamin Raskin, Article II, Section 1, Clauses 2 and 3, Nat’l Const. Ctr., https://constitutioncenter.org/interactive-constitution/interpretation/article-ii/clauses/350 [https://perma.cc/3RA2-YWH6]. South Carolina continued this practice through the 1832 election,52×52. See 1832 Presidential General Election Results, U.S. Election Atlas, https://uselectionatlas.org/RESULTS/national.php?year=1832&off=0&f=1 [https://perma.cc/EW4C-7JBV]. and as late as 1876, the legislature of recently admitted Colorado chose its electors directly.53×53. See 1876 Presidential General Election Results, U.S. Election Atlas, https://uselectionatlas.org/RESULTS/national.php?year=1876&f=0 [https://perma.cc/C39X-U3DD]. As the 1876 election was decided, amid great controversy, by a single electoral vote, the Colorado legislature’s decision proved decisive.54×54. See id. But even as the Colorado legislature declined to hold a popular vote for President, it did, in line with Article I, section 2, hold a popular election to choose its first member of Congress.55×55. See Statistics for October 3rd, 1876 Elections, Jerry Kopel, http://www.jerrykopel.com/b/first-election-1876.htm [https://perma.cc/RHX6-W9GM].

While no state today chooses its electors directly, two states, Maine and Nebraska, do deviate from the standard winner-take-all system, instead allocating electoral votes to the winners of each congressional district.56×56. See Maine & Nebraska, FairVote, https://www.fairvote.org/maine_nebraska [https://perma.cc/GB9S-ZNY3]. Because only two electors are awarded to the winner of the statewide vote, it would be possible for a larger state employing this system to assign the majority of its electors to a candidate who did not win the statewide vote. Even putting that contingency aside, a state’s choice to distribute its vote this way can have a significant effect on the election’s national outcome.57×57. See Harry Enten, Under a New System, Clinton Could Have Won the Popular Vote by 5 Points and Still Lost, FiveThirtyEight (Jan. 31, 2017, 10:28 AM), https://fivethirtyeight.com/features/under-a-new-system-clinton-could-have-won-the-popular-vote-by-5-points-and-still-lost [https://perma.cc/L5YS-S3EN]. Thus, state legislatures have, since the Founding, exercised potentially outcome-determinative discretion in awarding electoral votes. And of course, the Electoral College itself makes it possible for presidential candidates to win election without a plurality of the national popular vote. These observations, along with the document’s plain text, undermine the contention that the Constitution protects the right of voters to have a free choice of presidential candidates.

II. Supreme Court Precedent

Prior to the mid-twentieth century, the Supreme Court did not often hear challenges to state election laws. One notable exception was the 1892 case McPherson v. Blacker. That case concerned Michigan’s practice of appointing electors based on votes in individual congressional districts, similar to the system currently used in Nebraska and Maine.58×58. See McPherson v. Blacker, 146 U.S. 1, 24 (1892). This comparison ignores the fact that a system of this type would have potentially been far less democratic in 1892 than are the current systems of Nebraska and Maine, as states were then allowed to draw congressional districts with unequal populations. Cf. Gray v. Sanders, 372 U.S. 368, 378–81 (1963) (ending the practice). The plaintiffs’ theory was that this system violated Article II’s requirement that the state appoint electors “because all [the state’s] citizens otherwise qualified are not permitted to vote for all the presidential electors.”59×59. McPherson, 146 U.S. at 25. In rejecting this argument, the Court looked to the Constitution’s text and to the historical practice of states, concluding that the Constitution “leaves it to the [state] legislature exclusively to define the method of” choosing electors.60×60. Id. at 27. The Court also rejected challenges based on the Fourteenth and Fifteenth Amendments, holding that those amendments, rather than creating a right to vote for President or any other office, only barred discriminatory application of the right to vote as it existed under state law.61×61. See id. at 38. In keeping with Article II, section 1, clause 4 of the Constitution, the Court did strike down a portion of the Michigan law fixing a date for the meeting of its electors, as that date conflicted with the one provided by federal law. See id. at 40–41.

In 1934, the Supreme Court upheld an early campaign finance law, in the process affirming the power of Congress to regulate presidential elections.62×62. See Burroughs v. United States, 290 U.S. 534, 545 (1934). The Court cited no authority and provided little elaboration, remarking only on the importance of presidential elections and asserting that a contrary decision would “deny to the nation in a vital particular the power of self protection.” Id. Overturning this decision would call into question federal laws like EIGA, but that issue is separate from federal judicial oversight of elections through Anderson-Burdick. But the Court did not establish the basis for judicial oversight of such elections until 1968, when it decided Williams v. Rhodes.63×63. 393 U.S. 23 (1968). In that case, an Ohio law required new political parties to obtain signatures equivalent to fifteen percent of the votes cast in the previous gubernatorial election in order to secure a place on the presidential ballot, but required the Democratic and Republican parties to obtain only ten percent of the same figure.64×64. Id. at 24–26. Independent candidates were completely barred from the presidential ballot. Id. at 26. In evaluating this law, the Court considered Article II, section 1, conceding that the section granted “extensive power” to the states but concluding that “granted powers are always subject to the limitation that they may not be exercised in a way that violates other specific provisions of the Constitution.”65×65. Id. at 29. The Court then turned to the First and Fourteenth Amendments, identifying two constitutional rights — “the right of individuals to associate for the advancement of political beliefs, and the right of qualified voters, regardless of their political persuasion, to cast their votes effectively” — burdened by the law.66×66. Id. at 30. Because the state could not justify the unequal burden it placed on the rights of voters outside the two major parties, the Court struck the law down.67×67. See id. at 31–35. Williams was notable in that it placed limits on the “exclusive[]”68×68. McPherson v. Blacker, 146 U.S. 1, 27 (1892). discretion of states recognized by McPherson.69×69. Williams, 393 U.S. at 34. The decision did examine the text of Article II,70×70. Id. at 26. but it rested on political rights developed in cases unrelated to elections.71×71. See id. at 30 n.6 (collecting cases). Most notably, the Supreme Court applied a form of strict scrutiny taken from NAACP v. Button, 371 U.S. 415 (1968), a case that, far from dealing directly with voting rights, concerned a Virginia attempt to sabotage antisegregation activism. See id. at 417–26; Williams, 393 U.S. at 31.

In 1983, the Court again dealt with a presidential election in Anderson v. Celebrezze. The case concerned the 1980 presidential campaign of independent candidate Representative John Anderson, who won 6.6% of the national popular vote and no electoral votes.72×72. Anderson v. Celebrezze, 460 U.S. 780, 782, 784 (1983). In Ohio, Representative Anderson was initially refused a place on the ballot because he did not meet a March 20 filing deadline for independent candidates — even though the deadline did not apply to candidates nominated by a party.73×73. See id. at 782–83. Writing for the majority, Justice Stevens decided the case without reference to Article II, section 1. Instead, citing to Williams and to several cases unrelated to presidential elections, Justice Stevens concluded that “challenges to specific provisions of a State’s election laws” require courts to balance the injury to the plaintiff’s First and Fourteenth Amendment rights against the interests asserted by the state.74×74. See id. at 789. After comparing the burdens of the Ohio law against the state’s asserted interests, the majority concluded that the statute did not pass muster.75×75. See id. at 806. Then–Justice Rehnquist dissented on behalf of three other Justices. He began his analysis with Article II, section 1, describing it as one of the few constitutional provisions to grant “express plenary power to the States.”76×76. Id. (Rehnquist, J., dissenting). He reasoned that “the Constitution does not require that a State allow any particular Presidential candidate to be on its ballot” and that ballot access laws should be upheld so long as they “are rational and allow nonparty candidates reasonable access to the general election ballot.”77×77. Id. at 808. As noted by both the majority78×78. See id. at 789 (majority opinion) (denying the possibility of any “litmus-paper test”). and the dissent,79×79. See id. at 823 (Rehnquist, J., dissenting) (observing that the majority “draws no line”). the Court in Anderson drew no bright lines to determine how far a state could go in regulating elections, presidential or otherwise.

The case would find its counterpart in the 1992 case, Burdick v. Takushi. The plaintiff in that case was a Hawaii voter who wished to vote for a write-in candidate for the Hawaii legislature, as well as in other future elections.80×80. See Burdick v. Takushi, 504 U.S. 428, 430 (1992). Neither the majority nor the dissent attached significance to the fact that the restriction could include either presidential elections or elections governed by different parts of the Constitution. Hawaii law, however, categorically barred write-in votes.81×81. See id. at 430–32. Expressly disavowing the automatic application of strict scrutiny to laws burdening the right to vote,82×82. Id. at 432. Justice White wrote for the majority that such scrutiny applied only to “severe” restrictions.83×83. Id. at 434 (quoting Norman v. Reed, 502 U.S. 279, 289 (1992)). Meanwhile, “reasonable, nondiscriminatory restrictions” were subject to a lower standard, such that states’ regulatory interests are “generally sufficient to justify” them.84×84. Id. (quoting Anderson, 460 U.S. at 788). The Court concluded that the burden of Hawaii’s regulations was slight.85×85. See id. at 439. Thus, Hawaii’s stated interests in preventing “unrestrained factionalism” and voter “raiding” were sufficient justification.86×86. Id. (quoting Munro v. Socialist Workers Party, 479 U.S. 189, 196 (1986); quoting Tashjian v. Republican Party of Conn., 479 U.S. 208, 219 (1986)). Meanwhile, Justice Kennedy dissented on behalf of three Justices, explicitly agreeing with the standard articulated by the majority but reaching the opposite result.87×87. See id. at 445–46 (Kennedy, J., dissenting).

Thus was born the “now-familiar Anderson-Burdick balancing test.”88×88. Utah Republican Party v. Cox, 885 F.3d 1219, 1228–29 (10th Cir. 2018). The Court’s next major articulation came in the 2008 case Crawford v. Marion County Election Board.89×89. 553 U.S. 181 (2008). That case concerned an Indiana voter ID law that required identification to vote in person.90×90. Id. at 185 (plurality opinion). The Court upheld the law, but fractured in doing so. Justice Stevens wrote for a three-Justice plurality. His opinion cast Burdick as merely an application of the Anderson framework, and he stressed that framework’s flexibility.91×91. See id. at 190–91. Under his approach, Anderson-Burdick applied to any limitation on voters or political parties, “[h]owever slight,” and required a “sufficiently weighty” justification.92×92. Id. at 191 (quoting Norman v. Reed, 502 U.S. 279, 288–89 (1992)). The plurality concluded that, though the law had partisan motivations, it was “supported by valid neutral justifications,”93×93. Id. at 204. including combatting voter fraud, that justified the burden it placed on voters.94×94. See id. at 191. Meanwhile, three dissenting Justices in two opinions applied similar balancing tests but reached the opposite result.95×95. See id. at 209 (Souter, J., dissenting); id. at 237 (Breyer, J., dissenting). Justice Scalia concurred on behalf of two other Justices. Notably, he disagreed with the plurality’s understanding of Anderson-Burdick;96×96. Id. at 204 (Scalia, J., concurring in the judgment). he wrote that Burdick was the controlling opinion because it had “forged Anderson’s amorphous ‘flexible standard’ into something resembling an administrable rule.”97×97. Id. at 205. That rule called for a “two-track approach”98×98. Id. in which “nonsevere, nondiscriminatory restrictions” were reviewed deferentially and “laws that severely restrict the right to vote” were subjected to strict scrutiny.99×99. Id. at 204. Justice Scalia placed Indiana’s restrictions in the former category and thus concurred in upholding them.100×100. See id. at 209.

The divisions of the Crawford Court demonstrate certain failings of the Anderson-Burdick test. First, the Supreme Court cannot agree on what it is, leaving lower courts without clear guidelines.101×101. See Edward B. Foley, Essay, Voting Rules and Constitutional Law, 81 Geo. Wash. L. Rev. 1836, 1854–59 (2013); see also Mays v. LaRose, 951 F.3d 775, 783 n.4 (6th Cir. 2020) (“It’s unclear whether the Supreme Court ever intended Anderson-Burdick to apply to Equal Protection claims.”); Frank v. Walker, 769 F.3d 494, 500 (7th Cir. 2014) (Williams, J., dissenting from denial of rehearing en banc) (accusing the court of misreading Crawford); League of United Latin Am. Citizens of Iowa v. Pate, 950 N.W.2d 204, 213 (Iowa 2020) (accusing the dissent of misapplying Anderson-Burdick). Perhaps more importantly, even if the Court could agree, both proposed approaches are fundamentally indeterminate. The plurality did not pretend to draw any bright line, and thus the six Justices who supported open-ended balancing reached two results in three opinions. Justice Scalia’s approach is only slightly better in that it leaves courts without guidance as to whether restrictions are “severe.”

III. An Alternative Approach

The application of Anderson-Burdick to presidential elections has serious shortcomings. First, applying those cases to elections of all types ignores a fundamental feature of the Constitution — that the President, unlike members of Congress, is elected by electors chosen by the states. Second, as a policy matter, presidential elections are perhaps the single most contentious feature of American government, meaning that they require clearer judicial rules than Anderson-Burdick can provide.

One solution is a return to the deferential framework that prevailed prior to the mid-twentieth century. Under that approach, the plenary authority of states recognized in Bush v. Gore102×102. 531 U.S. 98, 104 (2000) (per curiam) (“[T]he state legislature’s power to select the manner for appointing electors is plenary . . . .”). would be acknowledged to include not just the hypothetical authority to appoint electors directly, but also the authority to give their people a limited choice. This approach would still acknowledge limits on state authority required by the Constitution, including protections of the right to vote — but it would not create an open-ended balancing test.

Recognizing the authority of states would reflect the basic interpretive rule that “the greater includes the lesser.” For instance, Congress has the power not to create the lower federal courts, and that power implies the ability to create courts of less than maximum jurisdiction.103×103. See Sheldon v. Sill, 49 U.S. (8 How.) 441, 448–49 (1850). The application of that proposition to presidential elections is straightforward: because states need not hold popular elections, they may therefore hold those elections in a limited way.104×104. This insight formed a large part of the Court’s reasoning in McPherson. See McPherson v. Blacker, 146 U.S. 1, 25–27 (1892). Just as that power includes the ability to hold presidential elections based on geographic divisions, it could also include the ability to hold elections without certain candidates or with stringent voting requirements. In addition to its logical force, this principle allows for compromise. For instance, imagine a state so concerned about voter fraud that it believed conducting an honest presidential election would be impossible under rules that would survive Anderson-Burdick.105×105. This is not wholly unrealistic — intense concern about voter fraud is increasingly mainstream. See, e.g., Howard Fischer, Proposed Law Would Allow Arizona Legislature to Overturn Presidential Election Results, Tucson.com (Feb. 13, 2021), https://tucson.com/news/state-and-regional/proposed-law-would-allow-arizona-legislature-to-overturn-presidential-election-results/article_c2a70681-59c0-512f-ba86-2bf23128f9ee.html [https://perma.cc/2FMW-J8J9]. Giving that state a freer hand would discourage it from choosing to not hold an election at all.

A recognition of state authority would also align with the structural values that are served by the United States’ unusual method of selecting leaders. One of those values is federalism. The formula for allocating electors — the number of representatives plus the number of senators — gives greater power to small states than would a system based on popular vote. According to James Madison, the system was intended in part as a compromise between small and large states.106×106. See Letter from James Madison to Henry Lee (Jan. 14, 1825), in 3 The Records of the Federal Convention of 1787, 464, 464 (Max Farrand ed., 1911). Even if one is not an originalist, this same argument is regularly made in favor of the Electoral College today.107×107. See, e.g., John Hendrickson & Stephen M. King, Federalsim [sic] Depends on the Electoral College, The Gazette (Dec. 26, 2019, 6:00 AM), https://www.thegazette.com/guest-columnists/federalsim-depends-on-the-electoral-college [https://perma.cc/WRK5-EG5M]. Indeed, the selection of presidential electors is one of the few constitutional prerogatives that is explicitly delegated to the states — there is thus a textual basis for concluding that it is part of that amorphous “substantial portion of the Nation’s primary sovereignty” that is reserved to the states.108×108. Alden v. Maine, 527 U.S. 706, 714 (1999). All of this stands in striking contrast to the election of members of the “People’s House,”109×109. See, e.g., The People’s House?, FairVote (July 18, 2007), https://www.fairvote.org/the_people_s_house [https://perma.cc/AJP8-E822]. which the Constitution specifies shall be chosen by “the People of the several States,”110×110. U.S. Const. art. I, § 2, cl. 1. or the power to choose senators, which was specifically stripped from the states in order to promote greater democracy.111×111. See id. amend. XVII; Ralph A. Rossum, The Irony of Constitutional Democracy: Federalism, the Supreme Court, and the Seventeenth Amendment, 36 San Diego L. Rev. 671, 672 (1999). The Electoral College is the last national political body that is still legally chosen by states as states, and for that reservation of authority to be meaningful, courts must recognize it as including a degree of discretion.

Another value underpinning the Electoral College is, put bluntly, elitism. In extolling the virtues of the Electoral College, Alexander Hamilton noted that it would place the selection of the President in the hands of “[a] small number of persons, selected by their fellow-citizens from the general mass.”112×112. The Federalist No. 68, at 410 (Alexander Hamilton) (Clinton Rossiter ed., 1961). Indeed, because the electors cannot be federal officials,113×113. See U.S. Const. art II, § 1, cl. 2. one original purpose of the Electoral College may have been to ensure that the power to select the President was vested in a group independent of both the government and the broader population.114×114. See Luis Fuentes-Rohwer & Guy-Uriel Charles, The Electoral College, the Right to Vote, and Our Federalism: A Comment on a Lasting Institution, 29 Fla. St. U. L. Rev. 879, 887–89 (2001). While less common than arguments about federalism, the idea that electors should exercise their own discretion in choosing the President has not vanished from popular discourse.115×115. See, e.g., Lawrence Lessig, The Constitution Lets the Electoral College Choose the Winner. They Should Choose Clinton., Wash. Post (Nov. 24, 2016), https://www.washingtonpost.com/opinions/the-constitution-lets-the-electoral-college-choose-the-winner-they-should-choose-clinton/2016/11/24/0f431828-b0f7-11e6-8616-52b15787add0_story.html [https://perma.cc/NK49-SAFM]. The Supreme Court has declined to recognize the rights of electors to make their own decisions independently of the states that selected them.116×116. See Chiafalo v. Washington, 140 S. Ct. 2316, 2320 (2020). The decision in Chiafalo v. Washington, 140 S. Ct. 2316, was also notable in that it rested on both “text [and] history,” id. at 2328, and some commentators have argued that the importance of history in its analysis serves to “entrench a particular and modern view of political participation,” Guy-Uriel E. Charles & Luis E. Fuentes-Rohwer, Chiafalo: Constitutionalizing Historical Gloss in Law & Democratic Politics, 15 Harv. L. & Pol’y Rev. 15, 19 (2020). Time will tell how influential this aspect of Chiafalo proves, though even an inquiry grounded in recent history would leave states with substantial power to place new limits on elections. Cf. Rachel Reed, “In Many Parts of the Country, the Voting Rights Act” Is “Close to a Dead Letter,” Harv. L. Today (July 8, 2021), https://today.law.harvard.edu/in-many-parts-of-the-country-the-voting-rights-act-is-close-to-a-dead-letter [https://perma.cc/SK2V-QY7M] (observing that even the use of a 1982 benchmark for voting rights would offer no protection against anything but “draconian” restrictions). But that conclusion still allows electoral votes to be awarded based on the wishes of the state political elite. The elitist function of the Electoral College counsels in favor of deference to state authority over the voting masses.

It is also worth briefly reiterating one of the values that does not underpin the Electoral College — that of national democracy.117×117. The Electoral College may provide some protection to democracy within a state, but only insofar as it vests the selection of electors in democratically elected state legislatures. As noted previously, many states at the time of the Founding did not even consult their people before choosing electors.118×118. See Ceaser & Raskin, supra note 51. Though that is no longer the norm, the Electoral College has allowed Presidents to be elected without a popular plurality as recently as 2016. This reality is sufficiently entrenched that the Supreme Court has never considered extending the “one person, one vote” rule to presidential elections.119×119. See Matthew J. Festa, Note, The Origins and Constitutionality of State Unit Voting in the Electoral College, 54 Vand. L. Rev. 2099, 2137 (2001). Meanwhile, the peculiarities of a single state, be that Colorado’s 1876 decision not to hold a presidential election or Florida’s 2000 decision to allow counties to choose “butterfly ballots,” can and do swing elections.120×120. See Spenser Mestel, How Bad Ballot Design Can Sway the Result of an Election, The Guardian (Nov. 19, 2019, 2:00 AM), https://www.theguardian.com/us-news/2019/nov/19/bad-ballot-design-2020-democracy-america [https://perma.cc/6293-8R7K]. It is thus misleading to say, as Anderson does, that any state election restriction “places a significant state-imposed restriction on a nationwide electoral process.”121×121. Anderson v. Celebrezze, 460 U.S. 780, 795 (1983). American presidential elections are nationwide only in the sense that they happen in every state at once. Acknowledging this reality makes clear that leaving states discretion to manage their own elections contravenes no constitutional principle.

Turning to practical considerations, elections generally are contentious subjects, and presidential elections especially so. While many areas of constitutional law are indeterminate, that weakness is especially problematic in such a politically fraught context. This concern drives the doctrine of Purcell v. Gonzalez,122×122. 549 U.S. 1 (2006) (per curiam). which, as more recently summarized, holds that “lower federal courts should ordinarily not alter the election rules on the eve of an election.”123×123. Republican Nat’l Comm. v. Democratic Nat’l Comm., 140 S. Ct. 1205, 1207 (2020) (per curiam). For elections to function smoothly, the rules must be known to all parties in advance.124×124. Cf. Purcell, 549 U.S. at 4–5 (“Court orders affecting elections, especially conflicting orders, can themselves result in voter confusion and consequent incentive to remain away from the polls. As an election draws closer, that risk will increase.”). And whenever those rules are determined, the passions aroused by presidential elections seem to influence courts. For instance, the aptly named Republican National Committee v. Democratic National Committee,125×125. 140 S. Ct. 1205. perhaps the most high-profile Supreme Court case of the 2020 election, produced a 5–4, party-line vote, despite the fact that it concerned the seemingly technical question of whether a state’s extension of its mail-in vote receipt deadline also extended the postmark deadline.126×126. See id. at 1206. This case was not decided on the basis of Anderson-Burdick, but the framework has directly produced its own share of questionable recent decisions, particularly in light of the Supreme Court’s muddled decision in Crawford. The Sixth Circuit’s decision in Obama for America v. Husted, 697 F.3d 423 (6th Cir. 2012), struck down a Republican-implemented voting restriction but established no clear principle, see id. at 425; Foley, supra note 101, at 1839. Eight years later, the same court applied Anderson-Burdick to a similar voting restriction in the same state but reached the opposite result. See Mays v. LaRose, 951 F.3d 775, 784–91 (6th Cir. 2020). Anderson-Burdick, meanwhile, hinges on judgments of the relative value of voting rights and election regulations. But America’s parties differ dramatically on that subject,127×127. For instance, Republicans and Democrats differ on the relative importance of expansive voting rights and effective voting regulations. See, e.g., Domenico Montanaro, Poll: More Americans Are Concerned About Voting Access than Fraud Prevention, NPR (July 2, 2021, 5:00 AM), https://www.npr.org/2021/07/02/1012302107/poll-more-americans-are-concerned-about-voting-access-than-fraud-prevention [https://perma.cc/ZV7Y-CXMY]. and there is no indication that courts can rise above this partisan divide and adjudicate cases in a principled way. A hands-off approach navigates these pitfalls by leaving partisan bickering in the political branches where it is more at home.128×128. Most of these challenges derive from the extent to which elections arouse popular passions, and presidential elections especially arouse those passions. For instance, midterm elections, in which the President is not on the ballot, typically generate about two-thirds the turnout of general elections. See Voter Turnout, FairVote, https://www.fairvote.org/voter_turnout#voter_turnout_101 [https://perma.cc/GA3S-VLSQ].

It is also worth acknowledging that, beyond issues of workability and constitutional fidelity, something is lost by an approach that constrains the ability of states to act as laboratories of democracy. Put differently, maybe states should require presidential candidates to release their tax returns. The release of tax returns serves important functions, like giving voters insights into a candidate’s conflicts of interest.129×129. See Danielle Lang, Candidate Disclosure and Ballot Access Bills: Novel Questions on Voting and Disclosure, 65 UCLA L. Rev. Discourse 46, 48 (2017). Perhaps for that reason, as of 2017, large majorities in both parties wanted President Trump to release his tax returns.130×130. See Justin Green, 64% of Republicans Want to See Trump’s Tax Returns, Axios (Apr. 13, 2017), https://www.axios.com/64-of-republicans-want-to-see-trumps-tax-returns-1513301586-84a1e448-f844-4ff5-87f9-c6c6bae3fbb4.html [https://perma.cc/Y7Q6-AFNT]. And laws like California’s S.B. 27 can be effective in forcing greater disclosure — the bill remains in force for gubernatorial elections, and during the state’s 2021 recall election, observers credited it with discouraging nonserious candidates.131×131. See Laurel Rosenhall, Will Tax Return Rule Scare Off Newsom Recall Candidates?, CalMatters (June 30, 2021), https://calmatters.org/politics/2021/06/newsom-recall-tax-returns [https://perma.cc/2955-6R6N]. But as long as President Trump, or other candidates like him, remain politically viable, it will be impossible for any state to implement a similar rule for presidential elections without severely burdening the ability of its citizens to choose from available candidates.132×132. This reality also illustrates the limitations of a judicial inquiry that strikes down election laws if they are rooted in improper partisanship. Cf. Foley, supra note 101, at 1860–64 (proposing such an inquiry). And while the merits of other election regulations differ, it is plausible that state governments, rather than federal courts, are best equipped to weigh those restrictions’ costs and benefits.

Although the approach described above may seem to go too far toward judicial abdication, there is a case to be made that it should go even further — the selection of presidential electors could be deemed a nonjusticiable political question.133×133. The Supreme Court’s opinion in Williams v. Rhodes dismissed this argument, 393 U.S. 23, 28 (1968), claiming that it had been rejected in both McPherson and Baker v. Carr, 369 U.S. 186 (1962). Both claims are questionable. McPherson did reject the argument that “all questions connected with the election of a presidential elector are political in their nature,” McPherson v. Blacker, 146 U.S. 1, 23 (1892), but went on to recognize the states as having “plenary power,” id. at 35. This conclusion might be translated into a determination of nonjusticiability in light of the doctrine’s subsequent evolution. See Tara Leigh Grove, The Lost History of the Political Question Doctrine, 90 N.Y.U. L. Rev. 1908, 1911 (2015) (concluding that, historically, the political question doctrine applied where factual determinations were better left to a political branch). Baker, of course, was a seminal case in defining the modern political question doctrine, but it dealt with malapportionment of congressional districts rather than presidential elections. See Baker, 369 U.S. at 187–88. Baker v. Carr134×134. 369 U.S. 186. set out a six-factor test for identifying political questions, but that test has since been distilled by Nixon v. United States135×135. 506 U.S. 224 (1993). and subsequent cases to focus on two factors: “a textually demonstrable constitutional commitment of the issue to a coordinate political department; or a lack of judicially discoverable and manageable standards for resolving it.”136×136. Id. at 228 (quoting Baker, 369 U.S. at 217). Though the others continue to exist, the six Baker factors “are probably listed in descending order of both importance and certainty.” Vieth v. Jubelirer, 541 U.S. 267, 278 (2004) (plurality opinion). Thus, the earlier factors carry greater weight. Notably, the third Baker factor, “the impossibility of deciding without an initial policy determination of a kind clearly for nonjudicial discretion,” Baker, 369 U.S. at 217, seems also to weigh in favor of considering presidential election regulation to be a political question because it requires initial policy judgments that the Constitution delegates to the states. Although only one factor is needed,137×137. See Rucho v. Common Cause, 139 S. Ct. 2484, 2500–02, 2506–07 (2019) (deeming partisan redistricting a political question solely on the basis that it is not subject to judicially manageable standards). both exist with regard to the selection of presidential electors. The relevant language of Article II, “in such Manner as the Legislature thereof may direct,”138×138. U.S. Const. art. II, § 1, cl. 2. is, in context, at least as strong as the language, “[t]he Senate shall have the sole Power to try all Impeachments,”139×139. U.S. Const. art. I, § 3, cl. 6. that led the Nixon Court to find a political question.140×140. See Nixon, 506 U.S. at 229. Though the clause at issue in Nixon, unlike that in Article II, includes the word “sole,” in context that word can be read to signify only that the Senate holds the power rather than the House of Representatives. See id. at 241–42 (White, J., concurring in the judgment). Meanwhile, the Supreme Court has declined to find that the power to regulate congressional districts is textually committed to states, see Rucho, 139 S. Ct. at 2495–96, but that power, unlike the power to appoint presidential electors, is subject to congressional alteration, U.S. Const. art. I, § 4, cl. 1. As for whether presidential election disputes can be resolved by “judicially discoverable and manageable standards,” that inquiry is famously unclear.141×141. See Kimberly Breedon, Remedial Problems at the Intersection of the Political Question Doctrine, the Standing Doctrine, and the Doctrine of Equitable Discretion, 34 Ohio N.U. L. Rev. 523, 536 (2008). However, scholars have teased out a series of factors for judging the manageability of a standard, including that standard’s intelligibility and practicability, as well as whether the subject matter makes a standard’s uncertainty worthwhile.142×142. See Richard H. Fallon, Jr., Judicially Manageable Standards and Constitutional Meaning, 119 Harv. L. Rev. 1274, 1285–96 (2006). Anderson-Burdick represents the Court’s best effort to craft a judicially manageable standard for elections, and, as noted above, it is wanting in many ways.

To be clear, this Note does not argue that it is necessary for courts to deem presidential elections political questions. For one thing, it is not clear that the “coordinate political department” of the first Baker factor may be a state.143×143. See Baker v. Carr, 369 U.S. 186, 217 (1962); id. at 210 (“[I]t is the relationship between the judiciary and the coordinate branches of the Federal Government, and not the federal judiciary’s relationship to the States, which gives rise to the ‘political question.’”). However, the neat applicability of the doctrine suggests that a large measure of deference is in order. The McPherson approach, in which states have plenary authority over the choice of electors unless they act contrary to another constitutional provision,144×144. See McPherson v. Blacker, 146 U.S. 1, 35 (1892). strikes this balance.

Regardless of how the Supreme Court replaces Anderson-Burdick, it will have to grapple with the issue of stare decisis. Though the criteria for overturning constitutional precedent are not always clear,145×145. See William Baude, Precedent and Discretion, 2019 Sup. Ct. Rev. 313, 329–30. in one formulation the Court considers whether the past decision was “egregiously wrong,”146×146. Ramos v. Louisiana, 140 S. Ct. 1390, 1414 (2020) (Kavanaugh, J., concurring in part). whether the decision has “caused significant negative jurisprudential or real-world consequences,” and whether overruling the decision would “unduly upset reliance interests.”147×147. Id. at 1415. All three considerations weigh in favor of overruling Anderson and Burdick.148×148. In addition, doing so would extend the modern trend away from the use of balancing tests in First Amendment law. See Kyle Langvardt, Can the First Amendment Scale?, 1 J. Free Speech L. 273, 278 (2021). The framework is wrong in that it is divorced from constitutional text and structure. It has also caused the judicial morass described above, while failing to provide meaningful protection to voters.149×149. See Joshua A. Douglas, Undue Deference to States in the 2020 Election Litigation, 30 Wm. & Mary Bill Rts. J. (forthcoming 2021) (manuscript at 1) (on file with the Harvard Law School Library). As for reliance interests, the elections that occurred under Anderson-Burdick are now accomplished facts, making it hard to see who would be worse off for having previously relied on the framework once it is overturned.150×150. Cf. Vikram David Amar, Justice Kagan’s Unusual and Dubious Approach to “Reliance” Interests Relating to Stare Decisis, Justia: Verdict (June 1, 2021), https://verdict.justia.com/2021/06/01/justice-kagans-unusual-and-dubious-approach-to-reliance-interests-relating-to-stare-decisis [https://perma.cc/HET2-TQ8M]. Thus, stare decisis poses no insurmountable obstacle to overruling Anderson and Burdick.

The transition away from Anderson-Burdick would be further eased by the fact that, especially during the 2020 election cycle, many courts have applied the cases in a limited way. During that election, states confronted with the COVID-19 pandemic made last-minute changes to their voting rules.151×151. See, e.g., Aria Jones & Sami Sparber, How Texas Has Made It Easier and Harder for People to Vote in the Pandemic, Tex. Trib. (Oct. 8, 2020, 5:00 AM), https://www.texastribune.org/2020/10/08/texas-voting-laws-coronavirus-pandemic [https://perma.cc/3TVC-NRR3]. Federal courts showed great deference, both in upholding measures making voting easier and in declining to intervene when states failed to implement such measures.152×152. See Douglas, supra note 149 (manuscript at 5, 11). Courts also showed deference after the election, when they declined to overturn the results of elections conducted by states.153×153. See id. (manuscript at 2). With Anderson-Burdick already waning in practice, abandoning it would be a relatively painless step.

IV. Remaining Limits

Anderson-Burdick is in theory derived from the First and Fourteenth Amendments,154×154. See Anderson v. Celebrezze, 460 U.S. 780, 789 (1983). and it has been argued that the First Amendment’s guarantee of freedom of association can, in fact, justify much of the caselaw.155×155. See Daniel P. Tokaji, Voting Is Association, 43 Fla. St. U. L. Rev. 763, 763, 765–66 (2016). But with the background presumption that states have authority to choose electors, or determine the means for choosing electors, as they see fit, it is impossible to justify such a wide-ranging and amorphous standard for presidential elections. Instead, a better approach would look to specific constitutional provisions to limit the broad authority of states — and it would remain cognizant of the distinction between presidential and other elections. This Part briefly sketches what some remaining constitutional limitations might be.156×156. Congress also has the power to regulate presidential elections, but see supra note 62 and accompanying text, and does so through legislation like EIGA and the National Voter Registration Act, 52 U.S.C. §§ 20501–20511.

First, it is worth clarifying that, while there is no right to ballot access,157×157. See Turner v. Fouche, 396 U.S. 346, 362 (1970) (discussing the right to vote in the context of the Equal Protection Clause, rather than as an absolute right). the Constitution does enshrine the right to vote, albeit in negative terms. The Fifteenth, Nineteenth, Twenty-Fourth, and Twenty-Sixth Amendments all bar various kinds of discrimination between voters,158×158. This fact counsels against reading into the Fourteenth or First Amendments a general right to vote that would render these other amendments redundant. and courts from McPherson onward have had no trouble concluding that these protections apply once a state has decided to entrust its choice of presidential electors to the voters.159×159. See McPherson v. Blacker, 146 U.S. 1, 38 (1892). While these amendments would likely not pose a major obstacle to restrictions like voter ID laws, they would prevent states from disenfranchising voters on the bases listed in the amendments.

The Equal Protection Clause of the Fourteenth Amendment also plays a role in protecting voters.160×160. See, e.g., id.; Bush v. Gore, 531 U.S. 98, 104 (2000) (per curiam). And because its terms are not limited to the casting of votes, it may well have a further role to play in elections, including presidential elections. The precise meaning of “equal protection of the laws”161×161. U.S. Const. amend XIV, § 1. is of course contested, but even under conservative readings, it includes “exemption from legal discriminations, implying inferiority in civil society.”162×162. Strauder v. West Virginia, 100 U.S. 303, 308 (1880). Thus, a state could not hinder political participation on the basis of a suspect classification.163×163. State classifications on the basis of alienage are suspect under the Equal Protection Clause, see Graham v. Richardson, 403 U.S. 365, 372 (1971), but the Constitution effectively requires states to discriminate on that basis with regard to presidential candidates, see U.S. Const. art. II, § 1, cl. 5 (“No Person except a natural born Citizen . . . .”). But that clause has not been read as informing equal protection analysis, and the adoption of a more deferential approach in this area need not change that. Rules designed to enhance participation by historically underrepresented groups might be a close case.164×164. Such a policy would be the electoral equivalent of affirmative action, a context in which strict scrutiny applies but is not “fatal in fact.” Adarand Constructors, Inc. v. Peña, 515 U.S. 200, 237 (1995) (quoting Fullilove v. Klutznick, 448 U.S. 448, 519 (1980) (Marshall, J., concurring in the judgment)). On the other hand, nonsuspect classifications, such as differing voting rules based on age,165×165. See Mass. Bd. of Ret. v. Murgia, 427 U.S. 307, 313–14 (1976) (per curiam). would probably be permissible under this clause. More generally, a deferential approach would not engage in the kind of broad, freewheeling equal protection analysis that has informed the Anderson-Burdick line of cases.166×166. See Williams v. Rhodes, 393 U.S. 23, 30 (1968) (“In determining whether or not a state law violates the Equal Protection Clause, we must consider the facts and circumstances behind the law, the interests which the State claims to be protecting, and the interests of those who are disadvantaged by the classification.”). The Court’s formulation in Williams is out of step with most modern equal protection jurisprudence, which extends strict scrutiny only to laws that draw suspect classifications. See, e.g., Kenji Yoshino, The New Equal Protection, 124 Harv. L. Rev. 747, 756 (2011).

Moreover, under U.S. Term Limits, Inc. v. Thornton,167×167. 514 U.S. 779 (1995). In that case, an amendment to the Arkansas Constitution barred ballot access for members of the U.S. House seeking a fourth term and for U.S. senators seeking a third. Id. at 783. Writing for the majority, Justice Stevens concluded that “with respect to Congress, the Framers intended the Constitution to establish fixed qualifications.” Id. at 792–93. While Justice Thomas’s argument to the contrary has merit, see id. at 845 (Thomas, J., dissenting), the case does not depend on Anderson-Burdick and would thus survive if that framework were overturned. the Constitution’s Qualifications Clauses would continue to function as ceilings on congressional ballot access restrictions, and the case’s reasoning likely extends to presidential elections as well.168×168. See Muller, supra note 10, at 87. States could not impose, say, a maximum age on presidential candidates. But, as explained by the Supreme Court, the holding only prevents states from “evading the requirements of the Qualifications Clauses by handicapping a class of candidates.”169×169. U.S. Term Limits, 514 U.S. at 831. Thus, the case might not be relevant to ballot access restrictions, like paperwork requirements, that are not analogous to the requirements of the Qualifications Clauses170×170. This reasoning is open to debate. Compare Chemerinsky, supra note 7 (concluding that S.B. 27 did not run afoul of the Qualifications Clause), with Griffin v. Padilla, 417 F. Supp. 3d 1291, 1302 (E.D. Cal. 2019) (concluding the opposite). or to requirements that singled out a candidate by name. And the case would place no barrier to more direct restrictions on the right to vote, such as voter ID laws.

Similarly, the First Amendment would continue to limit states. For instance, Establishment Clause jurisprudence, for all its famous incoherence,171×171. See, e.g., Paul Horwitz, The Agnostic Age: Law, Religion, and the Constitution 223 (2011). appears to ban state governments from endorsing a religion.172×172. See County of Allegheny v. ACLU, 492 U.S. 573, 592 (1989). A restriction limiting ballot access to one religious group would seem to run afoul of that bar, as would a restriction burdening participation by members of one or more religions. The Free Speech Clause, as well, would remain relevant. For instance, states may not engage in viewpoint discrimination,173×173. See R.A.V. v. City of St. Paul, 505 U.S. 377, 382 (1992). meaning they could not allow ballots or ballot access only to those who favored or opposed certain positions. And the freedom of expressive association would continue to protect various forms of engagement, including the formation and management of political parties174×174. See Cal. Democratic Party v. Jones, 530 U.S. 567, 574 (2000). — though it would not extend to the anonymous and individual act of casting a ballot for public office.

Finally, it bears emphasizing that constitutional limits do not have to come from the Federal Constitution. Just as S.B. 27 was struck down because it conflicted with the California Constitution,175×175. See Patterson v. Padilla, 451 P.3d 1171, 1191 (Cal. 2019). other election regulations must pass muster under the constitutions of states that pass them. The constitutions of all fifty states guarantee the right to vote,176×176. Joshua A. Douglas, To Protect the Right to Vote, Look to State Courts and State Constitutions, 9 Advance 21, 24 (2015). and leaving room for state courts to develop the implications of that right might make it stronger in the long term. Indeed, using a federal standard to “lockstep” this area of law “goes against the ideal of judicial federalism, which suggests that state constitutions should play a significant role in protecting individual liberties.”177×177. Id. at 29. Notably, this restraint is not compatible with the independent state legislature theory, under which state constitutions cannot limit state legislatures in their exercise of prerogatives granted by the federal constitution. See Morley, supra note 34, at 16.

All of these potential limitations, and others, raise questions of their own and would doubtless give rise to lawsuits. For instance, S.B. 27 would still face obstacles from the Qualifications Clause and federal preemption. Nonetheless, many of the sorts of issues that currently drive litigation — such as voter ID laws, signature requirements, or ballot-listing rules — would no longer have an obvious way into court.

Conclusion

There is no guarantee that states will use their power to fix the outcomes of presidential elections any more than they currently use their power to appoint electors directly. But despite that fact, and despite the safeguards described above, one cannot discount the possibility that America’s ever-degenerating partisan discord178×178. See generally Pew Rsch. Ctr., Political Polarization in the American Public (2014), https://www.pewresearch.org/politics/2014/06/12/political-polarization-in-the-american-public [https://perma.cc/E4YK-59ZK]. will lead to a future in which the majority of states, including some swing states, hold essentially one-candidate presidential elections.179×179. Non-major-party candidates might still be left on the ballot. Even if major-party candidates were omitted in only deep red or blue states, that would still mean the loss of a meaningful “national popular vote,” defeating one reform campaign. See Nat’l Popular Vote, https://www.nationalpopularvote.com [https://perma.cc/C8SQ-D4PB]. Yet even with no change in law, the Constitution does not require that Presidents be chosen by voters. Until the Electoral College is abolished, the power to elect Presidents rests ultimately in the states. This Note advocates merely for recognizing the implications of that reality. When passing on the constitutionality of presidential election restrictions, courts should cast aside the textually ungrounded Anderson-Burdick balancing test and instead defer to the states.